Anti piracy: dummy components etc to fool copiers

[Rich The Philosophizer]

On Fri, 12 Nov 2004 09:10:12 +1000, Adrian Jansen wrote:

budgie wrote:
On Thu, 11 Nov 2004 08:29:58 +1000, Adrian Jansen <adrian@qq.vv.net> wrote:


Release an even better version 1 week after the copy goes on the market.
and plan for even better versions. This both demoralises the copier,
and keeps you in a design job forever.


The voice of experience? ;-)

Yep ! Been designing stuff for 25 years now, and still not run out of
ideas.

Say, Adrian - do you have people who can take an invention to market?
Because I have several inventions that have been tripping over each other
in my head for quite some number of years, and none has found a market
yet - no, that's inaccurate. I _know_ the market is there - I just need
somebody who can sign an LC and _order_ a batch of the things:

The left-hand microwave oven.

The patent attorney I talked to verified my suspicion that it's not
patentable - refrigerators have had reversible doors for decades.

Now, mind you, I think it'd be foolish to try to market a microwave
oven with a _reversible_ door, but you could sell one with a door
that opens from left to right - how is your kitchen laid out? Fully
_half of the kitchens in the *world*_ that have a microwave oven
need one that opens from the other side.

Designing it is trivial. Mirror the cabinet.

And I so wish I'd had the facilities to take a video of my Mom getting
conked in the head by the door of the builtin in their apartment. With
a door that opens "the right way," the door would have opened up against
the wall.

But every single microwave oven in the _world_ has the door that
opens from right to left, from the user's POV.

Order a million of these puppies ( the ones where the door opens
left to right), and instantaneously capture 50% of the market of _all of
the microwave ovens sold in the entire *world*!_

So, if anybody out there can sign a PO for a million or so of these,
I'd appreciate a couple bucks spillage tossed my way, to serve as my
rags-to-riches Handsome Prince Benefactor Analogue. ;-)

Thanks,
Rich

 

[Mark Zenier]

In article <cmusr4$2d2$7@blue.rahul.net>,
Ken Smith <kensmith@green.rahul.net> wrote:

Label the parts with the wrong part numbers eg: get 74HC00s marked 74HC04

About $700, (last I saw), will buy a IC tester that will identify
a digital part.

Mark Zenier mzenier@eskimo.com Washington State resident

 

[ChrisGibboGibson]

Tom wrote:

I have developed an analogue circuit for my employer which works
unusually well. As we have several competitors who will take it apart to
see how it works, I now want to add some misdirection to prevent them
copying the circuit (it isn't novel enough to patent, but it outperforms
theirs by a significant margin). In fact I'd rather not copyright etc
the design anywhere as one of the groups I fear will copy it is a Far
Eastern bunch who would happily loot any published documentation.

I was wondering if anyone has any ideas about how to confuse copiers?

The circuit is low frequency, consumes about 50uA, surface mount, on a 4
layer board with a microcontroller on board. It's the analogue bit which
I want to protect.

You can't make it impossible to copy, but you can, of course, do things to make
it harder and harder.

Then, only someone really clever can copy it, and from his point of view it
would therefore, as he's so clever, be easier to just look at *what* yours
does, and design his own to do the same.

Obviously grind the part numbers off. If you can afford it, have the chips
printed with your own part numbers.

As it's low frequency, and you're using a 4 layer board, run signals all over
the place instead of directly to where they need to go, use hidden vias, and
where possible, bring the signal up *underneath* an IC. The copier's only
solution then is to physically check continuity between *every* connection on
the PCB (or remove the chips), all rather laborious and he may either make a
mistake or give in.

How much money can you afford to waste on it?

As you have a uC in there, you could phyisically add circuitry to do something
analogue (something totally bizarre and out of context - perhaps also using
some of your existing analogue circuitry as inputs) and feed the outputs back
to the uC, then completely ignore them in S/W

Alternatively, send signals *out* from the uC to mess up the analogue signals,
and ignore the analogue inputs during this time. That would make it rather
difficult to work out what is going on.

You can go on all day with tricks like that. But it will never stop someone who
is *totally* determined to copy it.

I would hazard a guess that only idiots would try to copy it in which case you
could fool them. But if they're stupid enough to be fooled by these tricks then
they probably wouldn't be able to copy it even if you left it as it is.

Someone clever enough to copy it is probably also clever enough to design his
own.

Gibbo

 

[Guy Macon]

Mark Zenier wrote:

About $700, (last I saw), will buy a IC tester that will identify
a digital part.

I own the company that produces those testers. They are sold out
and no longer in production, but we are a couple of months away
from introducing a new and improved model for $99. I will post
an announcement here when they become available.

 

[Ken Smith]

In article <cn2qaj$g5q$1@eskinews.eskimo.com>,
Mark Zenier <mzenier@eskimo.com> wrote:

In article <cmusr4$2d2$7@blue.rahul.net>,
Ken Smith <kensmith@green.rahul.net> wrote:

Label the parts with the wrong part numbers eg: get 74HC00s marked 74HC04

About $700, (last I saw), will buy a IC tester that will identify
a digital part.

They have to figure out that the label can't be trusted. That will delay
them. You can't prevent the copy but you can gain a little time.

--
--
kensmith@rahul.net forging knowledge

 

[john jardine]

"Ken Smith" <kensmith@green.rahul.net> wrote in message
news:cn3skk$rh0$1@blue.rahul.net...

In article <cn2qaj$g5q$1@eskinews.eskimo.com>,
Mark Zenier <mzenier@eskimo.com> wrote:
In article <cmusr4$2d2$7@blue.rahul.net>,
Ken Smith <kensmith@green.rahul.net> wrote:

Label the parts with the wrong part numbers eg: get 74HC00s marked
74HC04

About $700, (last I saw), will buy a IC tester that will identify
a digital part.

They have to figure out that the label can't be trusted. That will delay
them. You can't prevent the copy but you can gain a little time.

--
--
kensmith@rahul.net forging knowledge

A company I worked for bought out a competitor. Resources were pooled.

Flattered to discover that the competitor had previously spent 2 weeks
reverse engineering one of our products. I remember their drawings were more
accurate and useable than our originals.
regards
john

 

[Boris Mohar]

PCB is usually 0.063" thick.
0603 smd components are 0.06" long. Some are even round.
Your layout software must be able to call for unlpated thruhole pads.
Assembly can get tricky



Regards,

Boris Mohar

Got Knock? - see:
Viatrack Printed Circuit Designs http://www3.sympatico.ca/borism/

 

[Adrian Jansen]

As you found, inventions are easy. Getting them implemented and
marketed isnt. I have the same problem. I think Thomas Edison was
right, success is 1% inspiration and 99% perspiration.


--
Regards,

Adrian Jansen adrianjansen at internode dot on dot net
Design Engineer J & K Micro Systems
Microcomputer solutions for industrial control
Note reply address is invalid, convert address above to machine form.

 

[Rich The Philosophizer]

On Mon, 15 Nov 2004 08:21:59 +1000, Adrian Jansen wrote:

As you found, inventions are easy. Getting them implemented and
marketed isnt. I have the same problem. I think Thomas Edison was
right, success is 1% inspiration and 99% perspiration.

Don't forget "Having the Rockefellers put you on infinite expense
account."

;^j
Rich

 

[N. Thornton]

Rich The Philosophizer <null@example.net> wrote in message news:<pan.2004.11.12.05.53.54.695837@neodruid.org>...

Say, Adrian - do you have people who can take an invention to market?
Because I have several inventions that have been tripping over each other
in my head for quite some number of years, and none has found a market
yet - no, that's inaccurate. I _know_ the market is there - I just need
somebody who can sign an LC and _order_ a batch of the things:

The left-hand microwave oven.

this is a product idea, not an invention.

Designing it is trivial.

NT

 

[Rich Grise]

On Mon, 15 Nov 2004 19:28:44 -0800, N. Thornton wrote:

Rich The Philosophizer <null@example.net> wrote in message news:<pan.2004.11.12.05.53.54.695837@neodruid.org>...

Say, Adrian - do you have people who can take an invention to market?
Because I have several inventions that have been tripping over each other
in my head for quite some number of years, and none has found a market
yet - no, that's inaccurate. I _know_ the market is there - I just need
somebody who can sign an LC and _order_ a batch of the things:

The left-hand microwave oven.

this is a product idea, not an invention.

OK, that's fine. It doesn't matter what it's called. The point is, I can't
imagine it not selling millions of units, and I hope that whoever does the
first million lets some of their spillage fall my way.

Like I say, the design is trivial, and if I had the resources to
sign an LC for a hundred K or so, I'd just go ahead and do it.

Thanks,
Rich

 

[Terry Given]

Tom wrote:

Thank you guys,

an informative - and fun! - discussion. The thermite may not go down too
well with some customers, and we can't afford some of the solutions, but
there are some really good ideas there.

Just as we can't afford, say, a custom IC neither can our competitors
afford enormous amounts of time analysing the circuit, and though they
*might* extend to unsoldering the components and bussing it out; they
wouldn't have the resources to analyse the micro's code (though they
might *clone* it if they could). The market just doesn't support that
rich a manufacturing industry. So I shall use the simpler forms of
misdirection you suggest.

Our designs are checked by trusted 3rd parties then frozen, making
changes expensive. Product life cycles are several years. That rules out
frequent upgrades.

I also pondering Chip-on-board assembly. It is presumably a lot trickier
to analyse a die than a marked chip. Not sure our budget stretches to
that, depends what the contract manufacturer would add to the board cost.

Sorry, I cannot say what the circuit actually does.

Read the "the story of the P2" by R.A. Pease in the book edited by Jim
Williams - Analog Circuit Design. RAP discusses a not-so-parasitic
magnetic coupling between two adjacent circuit boards that was very
probably deliberate, and had the advantage of not only being non-obvious
(even on the schematic) but it required the thing be assembled right too.

A sneaky use of layout would quite likely be feasible on a multi-layer
pcb with external planes, hell, even on 2-sided pcb - many "designers"
dont seem to understand the layouyt forms an integral part of the
circuit, so probably wouldnt notice.

Mind you, years ago Williams brought out a game called (IIRC) stargate,
which was a mod to a Defender logic board. The brother of the guy who
fixed my folks power supplys and monitors reverse-engineered the
original board, which was potted and had a 40-pin DIP ULA of some
description. His equipment: chemicals, multimeter, 20MHz dual-trace
analogue CRO. Oh, and a very good neural net. He then built an add-on
pcb that did the same job, and for a lot less than Williams charged via
the only people who could import their stuff into NZ.



Cheers
Terry

 

[Terry Given]

Joel Kolstad wrote:

"Terry Given" <my_name@ieee.org> wrote in message
news:1vTkd.629$9A.37447@news.xtra.co.nz...

Mind you, years ago Williams brought out a game called (IIRC) stargate,
which was a mod to a Defender logic board. The brother of the guy who
fixed my folks power supplys and monitors reverse-engineered the original
board, which was potted and had a 40-pin DIP ULA of some description. His
equipment: chemicals, multimeter, 20MHz dual-trace analogue CRO. Oh, and a
very good neural net. He then built an add-on pcb that did the same job,
and for a lot less than Williams charged via the only people who could
import their stuff into NZ.


Sounds like a really savvy guy! Didn't the potted 'block' have to contain a
new (EP)ROM as well though? Or was that separate from the logic board mod?

Yeah the EPROMs needed to be re-programmed but that was easy. The potted
assembly plugged into a motherboard, along with the re-programmed ROMs
(that was a long time ago though, I can barely recall what they looked
like

And yep, he was a very smart cookie.

Cheers
Terry